Bivariate model-long run restriction

[marcusdixton]

Hi People,

I have to do exchange rate decomposition.
I have to use bivariate model and impose long run restriction like blanchard quah (1989).

The long run restriction is that in the long run the effect of nominal shock in real exchange rate is 0.

So basically I have nominal and real exchange rates data for a country and I have to see the impact of real shock in both the nominal and the real exchange rate. And then a nominal shock in both the real and exchange rate.

How do I impose the long run restriction? Could anyone help me?

I have eviews 11.

Thanks a lot!

[EViews Matt]

Hello,

I suggest you take a look at the Structural VAR section of the EViews documentation. A Blanchard-Quah style decomposition requires only a single zero restriction in the F matrix. Just keep in mind that you're performing a structural decomposition, so you're not producing shocks to your individual variables.